Download Should doctors recommend defibrillators for use at home after

head to head
Should doctors
recommend
defibrillators
for use at
home after
myocardial
infarction?
Michael F O’Rourke
thinks that enthusiasm
is an important factor;
Mathew Hutchinson and
David Callans say at home
defibrillators do not save
additional lives
Michael F O’Rourke professor of medicine, University of
New South Wales and St Vincent’s Clinic, 438 Victoria Street,
Darlinghurst, NSW 2010, Australia [email protected]
Ventricular fibrillation
causes most of the sudden
non-traumatic deaths in
Western society—up to 400 000 deaths a year in
the United States alone.1 Most deaths occur in
people without known heart disease, but many
occur in people with known coronary artery
disease.1 Defibrillators in aircraft, airports,
railway stations, sports stadiums, police patrol
cars, gymnasiums, and various public places
are designed for the population at large,2 and
in these locations their use is regarded as cost
effective.3 Like fire extinguishers, they seem to
be for use in a rare emergency.
Guidelines for the implantation of defibrillators now include patients of all ages whose
left ventricular ejection fraction is less than
30% but who have a prospect of living without major cardiac disability for two years.4
Implantation of a defibrillator is encouraged
by manufacturers and by proceduralists, who
fear that deviating from guidelines may be a
cause for legal action. Together with implanta-
DAVID GUYON, THE BOC GROUP PLC/SPL
Mathew D Hutchinson assistant professor of medicine,
University of Pennsylvania Health System,
Philadelphia, PA 19104, USA
David J Callans professor of medicine, University of
Pennsylvania Health System, Philadelphia, PA 19104, USA
[email protected]
All references are in the version on bmj.com
WHERE DO YOU STAND ON THE ISSUE?
Tell us on bmj.com
80
Few diseases have such disparate clinical outcomes in the
absence of definitive treatment
as sudden cardiac death. Many studies
have shown excellent outcomes in patients
resuscitated after sudden cardiac death who
receive early defibrillation, so it is logical
for doctors to seek to extend this powerful
treatment to other populations at high risk.
Unfortunately, the trials that have sought to
protect individual patients with automated
external defibrillators have been overwhelmingly unsuccessful. In light of the fortuitous
convergence of circumstances that leads to a
successfully resuscitated out of hospital cardiac arrest, the task of broad ­application of
this treatment seems indeed daunting.
First, defibrillation works only for patients
presenting with ventricular tachyarrhythmia. The current incidence of out of hospital cardiac arrest in the United States is
0.55 per 1000 people—about 160 000 events
a year.1 Of these arrests, only 20-38% will
have ventricular tachycardia or ventricular
fibrillation at presentation; thus most sudden cardiac arrests in the general popula-
tion, ­monitoring, and prior evaluation, they
cost about $100 000 in the United States during
the first year. Defibrillators were introduced
to save lives in patients with hearts “too good
to die,” but now they are used more often in
patients with hearts “too bad to live.”
Patients who have had a myocardial
infarction have a much higher risk of further
coronary events, including ventricular fibrillation, than other people of the same age.
Is it reasonable to suggest that such patients
have a defibrillator available in their home for
use by others in the event of cardiac arrest?1
Most such events do occur in the home, in the
bedroom or adjacent rooms.5 6 If such people
work in a federal US building or in a casino,
airport, or gymnasium or if they travel by air
in or towards the United States such a device
is available and a programme is in place for its
use. But not at home.
Economic argument
Those who argue against home defibrillators quote cost and the HAT (Home
­Automated External Defibrillator Trial) study.7
­Defibrillators can be purchased for around
tion are due to pulseless electrical activity,
bradyarrhythmia, or asystole.2 Although the
proportion of arrests caused by ventricular
tachycardia or ventricular fibrillation may be
higher in selected populations, the concept
is important. Furthermore, the overall incidence of ventricular tachyarrhythmia as a
cause of sudden cardiac death is decreasing:
undoubtedly a reflection of improvements
in the treatment of coronary heart disease.1
The proportion of sudden cardiac deaths
not attributable to ventricular tachycardia
or ventricular fibrillation is steadily increasing and has a uniformly poor outcome
even when the arrests occur in the hospital
with ready access to advanced life support
systems. The reduction in rates of sudden
death related to tachyarrhythmia is further
underscored by the lower than expected
event rates in trials of automated external
defibrillators (2.2% over 37 months from the
randomised, controlled Home Automated
External Defibrillator Trial (HAT)).3
Events need to be witnessed
Next, it seems intuitive that the successful resuscitation of sudden cardiac death
requires the event being witnessed. Unfortunately less than 20% of out of hospital arrests
occur in public locations. In the community
based, randomised, controlled Public Access
Defibrillation (PAD) trial, 95% of cases of
BMJ | 11 july 2009 | Volume 339
head to head
$1500, and people pay this much or more for
metallic paint or alloy wheels in their new car,
or for a better TV or sound system in their
home. The cost of purchasing a semiautomatic
defibrillator is around 2% that of implanting an
internal defibrillator, and in a heart “too good
to die” against a heart “too bad to live.” The
economic arguments do not make sense.
The HAT trial randomised 7001 patients
who had survived anterior myocardial
infarction a mean of 1.8 years previously and
who did not meet the criteria for implantation of an internal defibrillator.7 One group
was provided with a defibrillator at home
and the primary carer was instructed in its
use; the other group was not provided with
a defibrillator. Both groups had access to
community emergency care, which included
defibrillation. After average follow-up of 37
months there was no difference in outcome
of the two groups, and the primary end point
(death) was the same in each, around 2% a
year. The defibrillators that were used worked
perfectly, and there were no problems with
using them. I believe the HAT trial cannot be
used to exclude value of home defibrillation:
it shows the same problem as noted in an
earlier small trial conducted in Seattle.8
Participants in HAT (patients, carers, doctors) showed far less enthusiasm than investigators had anticipated. This necessitated heavy
promotion worldwide for entry of patients to
the trial, liberalisation of criteria to less sick
patients, and extension of the time from infarction to entry. Though it was not assessed, carers
were not enthusiastic about using the device in
elderly patients with cardiac and other comorbidities. Further, there was no organised followup to check the device or to check the training
of carers or assess their motivation. The device
was provided with “once-only” training; from
there on, the household was on its own. This
approach was justified at a national US meeting by a senior author who referred to a desire
to apply a “real world” evaluation. When the
trial finished, most devices were returned to
the hospital or centre of origin, even when
patients were still alive and at risk—resulting
in a sudden worldwide glut of second hand but
unused defibrillators, and reinforcing the view
that people did not wish to have the defibrillator for the family member.
Enthusiasm is important
What, then, of patients about to be discharged from hospital after a small or
medium myocardial infarction? I believe
that such patients and their spouses should
be advised to consider buying a defibrillator for the home, and that there should be
follow-up, such as did not occur in HAT, at
intervals of 6-12 months. Such a service is
arranged in Australia through the St John
Ambulance organisation.
Enthusiasm seems to be the key to successful defibrillation. Lack of enthusiasm
delayed the development of coronary care
wards and led to delay in installing defibrillators in ambulances. Lack of enthusiasm
led to withdrawal of defibrillators in some
police services. There is no point providing a
defibrillator to an unenthusiastic spouse, but
there is every reason for suggesting purchase
to an intelligent, enthusiastic spouse or carer,
provided there is appropriate follow-up.
sudden cardiac death who were dead on
arrival at the emergency department had
unwitnessed arrests.4 In HAT, only 36% of
the sudden deaths due to ventricular tachycardia or ventricular fibrillation occurring
at the patient’s own home were witnessed.
From this same study, neighbours or visitors
to the study patient’s home were twice as
likely to survive to hospital discharge from
a “shockable” rhythm (2/4, 50%) than the
study patients themselves (4/15, 27%). A
particularly disturbing finding in HAT was
the substantial number of cases of witnessed
arrests in which family members did not
even attach the defibrillator.
The most important factor for successful
resuscitation of ventricular tachycardia or
ventricular fibrillation is rapid defibrillation.
The efficacy of rapid defibrillation is best
assessed by time from onset of arrhythmia
to termination, but most events are not witnessed, so many trials index the time from
recognition of the arrest to defibrillation.
A recent observational study by Chan and
colleagues found that even in hospitalised
patients who were defibrillated within two
minutes of diagnosis, only 39% survived
to hospital discharge, and 40% had at least
moderate neurological disability.5 The biases
of in-hospital mortality studies notwithstanding, these results are both ­provocative and
disappointing.
The study which evaluated the best case
scenario for out of hospital resuscitation
was the observational, casino based defibrillation study by Valenzuela et al. This
study benefited from its highly public
settings, strategic placement of defibrillators (to minimise the time to first defibrillation), and operators trained in basic life
support and using defibrillators. In this
study, arrests were witnessed in 86% of
the patients and the mean time to the first
shock was 4.4 (SD 2.9) minutes. Patients
who were defibrillated within three minutes of their arrest had a 74% survival rate
to hospital discharge.6 To achieve such a
rapid and consistent response for the typically unwitnessed arrest in a patient’s home
seems unrealistic.
defibrillators, divergence of the survival
curves did not begin until nine months after
implantation.7 Furthermore, DINAMIT
(the randomised, controlled Defibrillator in
Acute Myocardial Infarction Trial) showed
no mortality benefit for implantable cardioverter defibrillators in patients immediately
after infarction. The high rate of non-arrhythmic death in the implantable defibrillator
arm of ­D INAMIT further suggests that
the dominant mechanism of mortality in
post-infarction patients with important left
ventricular dysfunction is heart failure, not
ventricular defibrillation.8
In summary, the incidence of arrhythmic
sudden death in post-infarction patients is
steadily decreasing. The prophylactic use of
defibrillators, either implanted or external,
have not been proven effective in reducing
overall mortality immediately post infarction. The high prevalence of unwitnessed
sudden cardiac death and the dramatic
reduction in patient survival with delayed
defibrillation makes the routine recommendation of home defibrillators impractical.
Further investigation is required to identify
higher risk subgroups of patients who will
benefit from implantation.
Mortality benefit?
Given the above discussion, it is not
­s urprising that HAT failed to show a
­mortality benefit with home use of defibrillators. At first glance the HAT participants
would seem to be an ideal population to
study, given their high prevalence of Q wave
anterior ­infarctions (64%) and mild left ventricle ­dysfunction. However, even in the higher
risk ischaemic population from MADIT II
(the randomised, controlled Second Multicenter Automated Defibrillator Implantation
Trial) who were treated with implantable
BMJ | 11 july 2009 | Volume 339 Competing interests: None declared.
Provenance and peer review: Commissioned; not
externally peer reviewed.
Cite this as: BMJ 2009;338:b874
Competing interests: None declared.
Provenance and peer review: Commissioned; not externally
peer reviewed.
Cite this as: BMJ 2009;338:b876
81